Heparin-Protamine Mismatch

Publisher Website
Google Scholar
Abstract
Objective: To test the effect of a new system designed to reduce heparin-protamine mismatch on bleeding after open heart surgery. Design: Nonrandomized but consecutive retrospective review of patients undergoing open heart surgery during a 9-month period. Setting: Multispecialty referral center. Patients: A total of 150 patients comparable by age, body surface area, and coagulation status undergoing primary open heart surgery for either coronary bypass or heart valve replacement. Intervention: In the first 75 patients (group 1), heparin sodium was neutralized with protamine sulfate, using a fixed ratio (1 mg of heparin sodium to 1.3 mg of protamine sulfate). An activated clotting time was used to confirm heparin neutralization. For the subsequent 75 patients (group 2), titration of heparin and protamine from defined lots was accomplished using activated clotting times adjusted and matched to drug lots to minimize biologic variability. Groups 1 and 2 had comparable operations, pump times, and cross-clamp times. Main Outcome Measures: Doses of heparin and protamine and their effect on blood product transfusion and postoperative bleeding were evaluated in all patients. Results: The average protamine sulfate dose for group 2 patients (287.56±8.3 mg) was significantly lower than that for group 1(346.01 ±12.6 mg) (P<.0005). Less protamine was associated with the transfusion of fewer red blood cells (0.92±0.15 vs 2.57±0.38 U) (P<.001), platelets (0.72±0.8 vs 2.96±0.80 U) (P<.01), and fresh-frozen plama (0.83±2.0 vs 2.01±0.48 U) (P<.03). No patients in group 2 required reexploration for bleeding, compared with eight patients in group 1. Conclusions: A reduction in protamine dose was associated with significant decreases in blood product use and postoperative bleeding. Excess protamine warrants consideration as both an important and a controllable factor in coagulopathy after open heart surgery. (Arch Surg. 1994;129:945-951)

This publication has 4 references indexed in Scilit: